1370 J. Agric. Food Chem.

2005, 53, 1370−1373

Rapid Determination of Polyphenols and Vitamin C in

Plant-Derived Products
STEPHANE GEORGEÄ ,*,† PIERRE BRAT,‡ PASCALINE ALTER,‡ AND MARIE J. AMIOT§
Centre Technique de la Conservation et des Produits Agricoles (CTCPA), Site Agroparc,
F-84911 Avignon Cedex 9, Centre de Coopération Internationale en Recherche Agronomique pour le
Développement (CIRAD), Département FLHOR, TA50/16, F-34398 Montpellier Cedex 5, and Human
Nutrition and Lipids, Faculté de Médecine, UMR U-476 INSERM (National Institute of Health and
Medical Research)/U-1260 INRA (National Institute for Agronomic Research),
27 Boulevard Jean Moulin, 13385 Marseille Cedex 05, France

Polyphenols, widely spread in our diet by the consumption of plant food products, are commonly
determined using Folin-Ciocalteu reagent that interacts with other different reducing nonphenolic
substances and leads to an overestimation of polyphenol content. In this paper we report an optimized
Folin-Ciocalteu method to specifically determine the contents of total polyphenols and vitamin C.
After the optimal conditions for the colorimetric assay were set, solid-phase extraction (Oasis HLB
(hydrophilic-lipophilic balance)) was carried out to eliminate the water-soluble reducing interferences
including vitamin C. Colorimetric correction was thus performed by subtracting interfering substances
contained in the water washing extract from the raw extract. Moreover, vitamin C present in the
water washing extract can be destroyed by heating and thus colorimetrically deduced. This procedure
was set up with synthetic solutions and validated on different extracts from fruit products.

KEYWORDS: Polyphenols; vitamin C; Folin-Ciocalteu; colorimetry; solid-phase purification; fruit; puree;

juice

INTRODUCTION In fact, polyphenols include other subclasses besides flavonoids,

During the past decade the interest in polyphenols, including
flavonoids, has considerably increased, especially by nutrition-
nists, epidemiologists, and food manufacturers. This is mainly
due to the discovery of various biological properties, namely,
their antioxidant effects and thereby their possible role in the
prevention of several chronic diseases involving oxidative stress
(1, 2). Polyphenols are the most abundant antioxidants in our
diet, since the average daily intake is about 1 g, which is almost
10-fold the intake of vitamin C, 100-fold the intake of vitamin
E, and 500-fold the intake of carotenoids (3). Foods and
beverages rich in polyphenols may have a large potential with
respect to prevention of diseases. This is the case for fruits and
vegetables associated with the prevention of stroke (4) and
cancers (5) and for tea (6) and soy that may protect breast cancer
(7). A food datababase on flavonoids, a large class of polyphe-
nols, was recently published by the USDA, http://www.nalus-
da.gov/fnic/foodcomp, based on the quality evaluation system
reported by Holden et al. (8). Such a database is extremely useful
for epidemiological studies on the relationship between dietary
flavonoids and health. However, it remains extremely difficult
to estimate the average content daily intake of total polyphenols.
such as phenolic acids, stilbenes, lignans, tannins, and oxidized
polyphenols. Many of these compounds display a large diversity
of structures and escape quantification, usually carried out by
HPLC and diode array detection (9), because (i) there is lack
of commercial standards and (ii) numerous structures are not
yet elucidated. In these cases, the total polyphenol content is
usually underestimated. Recently, Vinson et al. (10-12) reported
data on the total polyphenol content of various fruits and
vegetables. This content was colorimetrically measured by the
Folin-Ciocalteu reaction after correction for ascorbic acid
contribution. In fact, ascorbic acid is certainly the main reducing
agent, which can interfere in the Folin-Ciocalteu reaction.
However, other reducing substances such as some sugars and
amino acids could also interfere.
The aim of this study was to set up a rapid colorimetric
method to quantify total polyphenols in various fruits, veg-
etables, and derived products (purees, juices). The initial step
was the optimization of the colorimetric method using Folin-
Ciocalteu reagent. To quantify total polyphenols, solid-phase
extraction was performed. An additional step is proposed to
estimate vitamin C content. Results obtained with synthetic
solutions were validated with some foods and beverages.

* To whom correspondence should be addressed. Phone: +33(0) MATERIALS AND METHODS

490841709. Fax: +33(0) 490841726. E-mail: sgeorge@ctcpa.org.
† CTCPA. Sample Preparation. All of the food products (apple purees and
‡ CIRAD. juices) were purchased at a local market and stored at 4 °C until
§ UMR U-476 INSERM/U-1260 INRA. analysis.

10.1021/jf048396b CCC: $30.25 © 2005 American Chemical Society

Published on Web 01/28/2005
Total Polyphenols and Vitamin C Determination J. Agric. Food Chem., Vol. 53, No. 5, 2005 1371

Material and Solvent Standards. All solvents were of the highest Table 1. Colorimetric Response (Absorbance due to Folin−Ciocalteu
analytical grade. Reference compounds were from Extrasynthèse Reaction) before and after Methanol Elution of Different Polyphenol
(Genay, France) and reagents and solvents from Sigma-Aldrich Chimie Solutions Using the Oasis HLB Cartridge
(Saint Quentin Fallavier, France).
Cartridges. The SPE cartridge was an Oasis HLB from Waters phenolic quantity initial absorbance after percentage of no. of
(Milford, MA). HLB is an acronym for hydrophilic-lipophilic balance, compound (µg) absorbance methanol elution recovery assays
the solid-phase being a copolymer [poly(divinylbenzene-co-N-vinylpyr- gallic acid 420 0.383 0.348 91 4
rolidone)]. After any analysis, the cartridge was conditioned with 4 caffeic acid 330 0.236 0.222 94 3
mL of pure methanol and rinsed with 2 × 4 mL of water. chlorogenic acid 459 0.248 0.213 86 3
Preparation of Raw Extracts. Purees or juices (10-20 g) were catechin 345 0.362 0.279 77 4
polymer 183 0.141 0.091 65 4
homogenized with the same extraction solution (acetone/eau, 7/3, v/v) naringenin 390 0.087 0.087 99 3
for 30 min. Mixture supernatants were then recovered by filtration naringin 390 0.085 0.084 98 3
(Whatman, England) and constituted the raw extracts (REs). phloretin 954 0.493 0.351 71 3
Separation of Polyphenol and Other Water-Soluble Components phloridzin 486 0.174 0.154 88 3
by Solid-Phase Extraction. REs were added with distillated water to quercetin 462 0.394 0.315 80 3
reduce the proportion of acetone to 7%. Diluted REs (2 mL) were settled quercitrin 255 0.174 0.138 79 3
rutin 424 0.207 0.183 88 3
on an Oasis cartridge (Waters). Interfering water-soluble components luteolin 17 0.145 0.139 96 3
(reducing sugars, ascorbic acid) were recovered with 2 × 2 mL of luteolin 7-glucoside 34 0.045 0.048 100 3
distillated water. The recovered volume of the washing extract (WE) cyanidin 148 0.076 0.054 71 4
was carefully measured. cyanin 216 0.355 0.338 95 4
Elimination of Vitamin C from WE. Heating was carried out on
the washing extract (3 mL) for 2 h at 85 °C (Fisons Haake N2 oil
bath) and led to the heated washing extract (HWE). reflects the quantity of polyphenols usually expressed as gallic
Folin-Ciocalteu Assay. All extracts (RE, WE, and HWE) were acid equivalent (GAE) or catechin equivalent. The first step of
submitted to the Folin-Ciocalteu method (13), adapted, and opti- our study was to optimize the experimental conditions of this
mized: A 2.5 mL sample of water-diluted Folin-Ciolcateu reagent reaction: influence of the extraction solvent of polyphenols in
(1/10) was added to the different extracts. The mixture was incubated the colorimetric reaction mixture, reaction kinetics after addition
for 2 min at room temperature, and 2 mL of sodium carbonate (75 of the reagents (FC and sodium carbonate) under different
g‚L-1) was added. The mixture was incubated for 15 min at 50 °C and
temperatures. Generally, polyphenols are extracted with methanol/
finally cooled in a water-ice bath. The specific absorbance at 760 nm
was immediately measured. water or acetone/water (14). The proportion of the two organic
Determination of Total Polyphenols and Vitamin C and Expres- solvents, methanol and acetone, were tested in FC reaction.
sion of the Results. Total polyphenols, determined by subtracting gallic Dilution was necessary prior to the colorimetric reaction. In fact,
acid equivalent from RE from that of WE, were expressed as mg of the maximum organic solvent (methanol or acetone) percentage
gallic acid/100 g of product (slp ) 0.012, R2 ) 0.99). Linearity was was 7% without interference on the colorimetric reaction.
obtained between 50 and 500 mg/L corresponding to absorbance values Finally, all the extractions were performed with acetone/water
between 0.1 and 0.6. Vitamin C, determined by subtracting ascorbic (70/30, v/v), and then the mixture was water diluted 10-fold
acid equivalent from WE from that of HWE, was expressed as mg/ before colorimetric reaction. The kinetic reaction (from 2 to 20
100 g of product (slp ) 0.008, R2 ) 0.99). Linearity was obtained min) after addition of FC reagent (10-fold diluted in water) was
between 50 and 1000 mg/L corresponding to absorbance values between
tested. Two minutes at room temperature was fixed for the
0.1 and 0.6.
All the experiments were performed in triplicate. reaction time after addition of FC reagent. After addition of
HPLC Analysis. Vitamin C was quantified by HPLC (Waters sodium carbonate, the maximum stable color response was
system) using an isocratic gradient equipped with a reversed-phase C18 obtained for a 15 min reaction time at 50 °C (water bath). It
column (Waters, Spherisorb ODS 2) (5 µm packing) (250 × 4.6 mm could be stressed that a decline of the coloration started after
id). Ascorbic acid was eluted under the following conditions: injected 30 min.
volume 20 µL; oven temperature 30 °C; solvent mixture K2HPO4 (0.1 Results Expression. Under this optimized reaction, different
M), KH2PO4 (0.08 M), MeOH (55/25/20, v/v/v). The flow rate was standards (phenolic or not) were tested at different concentra-
1.5 mL min-1, and the total elution time was 10 min. Detection was tions. In reference to gallic acid, catechin, caffeic acid, narin-
performed by a III-1311 Milton Roy fluorimeter (Ivyland, PA) with genin, and rutin gave similar responses while glyscosides of
λexcitation ) 350 nm and λemission ) 430 nm. Quantification was carried
naringenin and hesperetin (naringin and hesperidin) displayed
out by external calibration with ascorbic acid. The calibration curve
was set from 1 to 7 µg/mL ascorbic acid. a lower absorbance (∼2-fold lower). Such behavior could lead
An HPLC instrument (Agilent 1100 Series) with diode array to an underestimation of these compounds. Vitamin C exhibited
detectors was used to check phenolic compounds in each extract a slightly lower response than gallic acid (∼80% of the gallic
according to the following method: column, 4.6 × 150 mm C18 (Alltima acid absorbance at the same concentration). By contrast,
5 µm, Alltech) with a C18 column guard with the same packing; carotenoids appeared to exhibit a higher response value (∼2-
polyphenols were eluted with a gradient of water (0.5% (v/v) formic or 3-fold higher than that of gallic acid). We can conclude that,
acid) and acetonitrile. In the first 10 min, acetonitrile was 10%, from in extracts rich in carotenoids, Folin-Ciocalteu could more or
10 to 35 min, acetonitrile increased from 10% to 50% and from 35 to less overestimate the polyphenol content. However, carotenoids
50 min acetonitile increased to 100%. The total elution time was 50 are poorly extracted with more polar solvents, suggesting their
min, and the flow rate was 1.0 mL min-1.
low contribution in the colorimetric assay.
Solid-Phase Extraction. Different cartridges were assayed
RESULTS AND DISCUSSION
as described by Antolovich and Robards (15), and only the Oasis
Folin-Ciocalteu Optimization. Folin-Ciocalteu reagent, a HLB gave exploitable results. Different polyphenols were
mixture of phosphotungstic (H3PW12O40) and phosphomolybdic tentatively eluted with methanol after their adsorption on the
(H3PMo12O40) acids, is reduced to blue oxides of tungstene solid phase of Oasis HLB (Table 1). The recovery percentage
(W8O23) and molybdene (Mo8O23) during phenol oxidation. This greatly varied according to the structures. For naringenin and
reaction occurring under alkaline conditions is carried out with its glycoside naringin, the recovery reached close to 100%. By
sodium carbonate. Blue coloration is followed at 760 nm and contrast, catechin was less eluted; 23% of catechin was
1372 J. Agric. Food Chem., Vol. 53, No. 5, 2005 Georgé et al.

Table 2. Total Polyphenols and Vitamin C Content as the Mean of

Three Replicates (mg (100 g of fresh weight)-1) in the Different
Products
food and total
beverage polyphenolsa vitamin Ca vitamin Cb vitamin Cc
apple puree 1 96.0 (97.7−94.3)d 29.3 (30.1−28.5) 40.9 (42.1−39.7) 43.5 (45.6−41.3)
apple puree 2 63.5 (65.2−62.0) 10.4 (11.8−9.0) 14.5 (15.9−12.9) 14.9 (15.5−14.3)
apple juice 57.1 (59−55.2) 24.4 (25.7−23.1) 34 (35.5−32.5) 37.2 (39.1−35.3)
orange juice 19.5 (20.9−18.1) 20.7 (22.1−19.3) 28.8 (30.8−27.0) 34.6 (35.9−33.2)
tomato juice 18.4 (20.1−16.7) 4.1 (5.8−2.4) 5.6 (7.9−3.3) 3.2 (3.4−3.0)

a Expressed in gallic acid equivalent (GAE) and determined by Folin−Ciocalteu

method. b Expressed in ascorbic acid equivalent and determined by Folin−Ciocalteu

method. c Expressed in ascorbic acid equivalent and determined by HPLC. d Values
in parentheses are the minimum and maximum values.

retained on the column. The retained percentage was greater

for polymerized catechins (procyanidins), 35%. In addition, other
flavonoids as aglycon forms (phloretin, quercetin, cyanidin)
appeared to be poorly eluted with methanol. Other elution
solvents (acetonitrile, methanol/methyl tributyl ether, dichlo-
romethane) were tested without more success. We conclude that
the normal protocol using the Oasis cartridge cannot be applied
for the complete desorption of total polyphenols, especially when
plant food products are particulary rich in tannins. In regard to
these results, the total polyphenol content was only assayed by
subtracting the interfering reducing substances from the Folin-
Ciocalteu response. Because in fruits and their derived products,
vitamin C is known to be the main interfering reducing
substance in the colorimetric assay, the efficiency of vitamin C Figure 1. Flowchart of the polyphenol and vitamin C determination
washing by water was checked using the Oasis cartridge. procedure (A, obtention of different extracts; B, Folin−Ciocalteu reaction).
Vitamin C was totally eluted with 2 × 2 mL of water. The WE
did not contain any polyphenol tested (caffeic and gallic acid,
cyanidin, and catechin), which was confirmed by HPLC. In a
manner similar to that of vitamin C, cysteine was totally eluted
during the washing step. In addition, Oasis HLB cartridges can
be used 5-fold after its regeneration with 3-fold of 2 mL of
methanol followed by 2-fold of 2 mL of water.
Colorimetric Assay of Vitamin C Using Folin-Ciocalteu.
As previously shown, ascorbate could interfere in the Folin-
Ciocalteu reaction. Its concentration cannot be directly measured
by the FC method on the washing extract, which could contain
other reducing substances. As reported by Vinson et al. (12),
vitamin C could be separately assayed by HPLC. However,
because vitamin C is known to be highly thermolabile, a thermal
treatment was chosen as an easier way for its determination.
By subtracting the FC response of the HWE from that of the
WE, we thus obtained the concentration of vitamin C. Optimal
conditions to eliminate the totality of vitamin C in synthetic
solutions were 2 h of heating at 85 °C. It could be noted that
no browning color was developed during such heating condi-
tions. To quantify total polyphenols and vitamin C, we propose
the scheme reported in Figure 1.
Application to Food Products and Beverages. Correction
by subtracting interfering substances appeared necessary to get
appropriate values for polyphenol determination. In fact, for
apple purees and juice, the contribution of the interfering
substances varied from 31% to 48% (Figure 2). This contribu- Figure 2. Total polyphenols and interferences in the different products
tion was higher for orange juice (77%) and tomato juice (70%). (mg (100 g of fresh weight)-1).
As expected, numerous reducing compounds could interfere in
the quantification of polyphenols by the FC method. Among low in tomato juice, vitamin C being only 9% of the total
them, vitamin C was supposed to have the major contribution. interferences expressed as GAE.
This was the case in one-apple derived product (enriched in In addition to the polyphenol quantification, the heating step
vitamin C). However, the contribution of vitamin C was 32% in the present colorimetric procedure could be proposed to
of the total interferences for orange juice and 46% for apple quantify vitamin C. This approach gave values which were in
juice (Table 2). The contribution of vitamin C was remarkably good agreement with those obtained by HPLC. However, it
Total Polyphenols and Vitamin C Determination
could be underlined that for low values of vitamin C an
overestimation was observed.
In conclusion, this study proposes a rapid colorimetric method
using Folin-Ciocalteu reagent to estimate the total polyphenol
content of foods and beverages, after removal of interfering
nonphenolic reductants. Even HPLC is a suitable method to
quantify each phenolic compound, it cannot correctly estimate
the total polyphenol content, as previously described in the
Introduction.
In the present method, food products require minimal sample
preparation before polyphenol analysis, by using specific solid-
phase extraction on minicartridges. In addition, vitamin C
quantification could be concomitantly performed in this pro-
cedure. The proposed method could be useful for a large panel
of applications (epidemiological studies, composition tables,
industry, ...). This approach could also be employed for human
fluids (plasmas and urines). Such studies are currently being
carried out.
ACKNOWLEDGMENT
We gratefully acknowledge the technical assistance of Victor
Pirisi from UMR U-476 INSERM/U-1260 INRA.
LITERATURE CITED
(1) Rice-Evans, C. Plant polyphenols: free radical scavengers or
chain-breaking antioxidants? Biochem. Soc. Symp. 1995, 61,
103-116.
(2) Lairon, D.; Amiot, M.-J. Flavonoids in food and natural
antioxidants in wine. Curr. Opin. Lipidol. 1999, 10, 23-28.
(3) Scalbert, A.; Williamson, G. Dietary intake and bioavailability
of polyphenols. J. Nutr. 2000, 130, 2073S-2085S.
(4) Ness, A. R.; Powles, J. W. Fruit and vegetables, and cardiovas-
cular disease: a review. Int. J. Epidemiol. 1997, 26, 1-13.
(5) Steinmetz, K. A.; Potter, J. D. Vegetables, fruit, and cancer
prevention: a review. J. Am. Diet. Assoc. 1996, 96, 1027-1039.
J. Agric. Food Chem., Vol. 53, No. 5, 2005 1373
(6) Jansen, M. C.; McKenna, D.; Bueno-De-Mesquita, H. B.;
Feskens, E. J.; Streppel, M. T.; Kok, F. J.; Kromhout, D.
Reports: Quantity and Variety of Fruit and Vegetable Consump-
tion and Cancer Risk. Nutr. Cancer 2004, 48, 142-148.
(7) Lamartiniere, C. A. Protection against breast cancer with
genistein: a component of soy. Am. J. Clin. Nutr. 2000, 71,
1705S-1709S.
(8) Holden, J. M.; Bhagwat, S. A.; Patterson, K. Y. Development
of a multi-nutrient data quality evaluation system. J. Food
Compos. Anal. 2002, 15, 339-348.
(9) Romani, A.; Minunni, M.; Mulinacci, N.; Pinelli, P.; Vincieri,
F. F.; Del Carlo, M.; Mascini, M. Comparison among differential
pulse voltammetry, amperometric biosensor, and HPLC/DAD
analysis for polyphenol determination. J. Agric. Food Chem.
2000, 48, 1197-1203.
(10) Vinson, J. A. Flavonoids in foods as in vitro and in vivo
antioxidants. AdV. Exp. Med. Biol. 1998, 439, 151-164.
(11) Vinson, J. A.; Su, X.; Zubik, L.; Bose, P. Phenol antioxidant
quantity and quality in foods: fruits. J. Agric. Food Chem. 2001,
49, 5315-5321.
(12) Vinson, J. A.; Proch, J.; Bose, P. Determination of quantity and
quality of polyphenol antioxidants in foods and beverages.
Methods Enzymol. 2001, 335, 103-114.
(13) Singleton, V. L.; Rossi, J. A. Colorimetry of total phenolics with
phosphomolybdic-phosphotungstic acid reagents. Am. J. Enol.
Vitic. 1965, 16, 144-158.
(14) Santos-Buelga, C.; Williamson, G. Methods in polyphenol
analysis; RSC Publications, Royal Society of Chemistry: Lon-
don, 2003.
(15) Antolovich, M.; Robards, K. Analytical chemistry of fruit
bioflavonoids a review. Analyst 1997, 122, 11R-34R.
Received for review September 24, 2004. Accepted December 14, 2004.
The present study was financially supported by the Ministère de la
Recherche et de la Technologie (France) and by the Regional Council
of Provence-Alpes-Cote d’Azur.
JF048396B